Automatic quantitative evaluation method for seismic data quality based on signal correlation

Abstract

This invention discloses an automatic quantitative evaluation method and system for the quality of seismic data during coal mining based on signal correlation, belonging to the field of seismic data processing technology during coal mining. The method includes: during coal mining, acquiring seismic signals from multiple locations within a preset time window using a sensor array, performing time alignment processing to obtain multiple seismic signal sequences; acquiring the working intensity parameters of the coal mining equipment, configuring a first anomaly classification standard based on the working intensity parameters; acquiring the influence amplitude of noise from the coal mining equipment within the seismic signals under the working intensity parameters, adjusting the first anomaly classification standard to obtain a second anomaly classification standard; using the second anomaly classification standard to classify the multiple seismic signal sequences during coal mining, obtaining anomaly coefficients, and calculating the seismic data quality coefficient as the evaluation result. This invention effectively improves the accuracy of coal mining geological data quality analysis.

Description

Technical Field

[0001] This invention relates to the field of seismic data processing technology during seismic mining, and specifically to a method and system for automatic quantitative evaluation of the quality of seismic data during seismic mining based on signal correlation. Background Technology

[0002] In coal mining, seismic monitoring during mining is a crucial means of ensuring mining safety, and high-quality seismic data is essential for assessing rock strata stability. However, seismic signals during mining are easily affected by noise from mining equipment, and fluctuations in equipment operating intensity can exacerbate the noise impact, leading to unstable data quality.

[0003] Existing seismic data quality assessment technologies typically rely on qualitative judgments based on characteristics such as waveform morphology, continuity, and background noise levels. While these technologies can support basic geological analysis in small-scale data scenarios, the resulting assessments lack objective and unified quantitative standards, leading to inaccuracies in coal mining geological data analysis. Furthermore, they are inadequate for handling massive, continuously generated monitoring data, hindering real-time assessment and feedback of data quality. Therefore, there is an urgent need for a method that can effectively improve the accuracy of coal mining geological data quality analysis. Summary of the Invention

[0004] This application provides a method and system for automatic quantitative evaluation of the quality of seismic data during mining based on signal correlation, which aims to solve the technical problem of inaccurate analysis of coal mining geological data in the prior art.

[0005] In view of the above problems, this application provides an automatic quantitative evaluation method and system for the quality of seismic data acquired during mining based on signal correlation.

[0006] Firstly, this application provides an automatic quantitative evaluation method for the quality of seismic data acquired during mining based on signal correlation, including: During the coal mining process, a sensor array is used to collect seismic signals from multiple locations within a preset time window, and time alignment processing is performed to obtain multiple seismic signal sequences. Obtain the working intensity parameters of the coal mining equipment, and configure a first anomaly classification standard based on the working intensity parameters; The influence amplitude of coal mining equipment noise under the working intensity parameter and the seismic signal is obtained. The first anomaly classification standard is adjusted to obtain the second anomaly classification standard. The second anomaly classification standard is used to classify the multiple seismic signal sequences obtained during mining, obtain anomaly coefficients, and calculate the seismic data quality coefficients obtained during mining as evaluation results.

[0007] Secondly, this application provides an automatic quantitative evaluation system for the quality of seismic data acquired during mining based on signal correlation, including: The signal acquisition and alignment module is used to acquire seismic signals from multiple locations within a preset time window during coal mining through a sensor array, and perform time alignment processing to obtain multiple seismic signal sequences. The intensity parameter configuration module is used to obtain the working intensity parameters of the coal mining equipment and configure the first anomaly classification standard according to the working intensity parameters; The noise standard correction module is used to obtain the influence amplitude of coal mining equipment noise under the working intensity parameter and the seismic signal during mining, and to adjust the first anomaly classification standard to obtain the second anomaly classification standard. The anomaly classification and evaluation module is used to classify the multiple seismic signal sequences obtained during mining using the second anomaly classification standard, obtain anomaly coefficients, and calculate the seismic data quality coefficients obtained during mining as evaluation results.

[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages: This application provides an automatic quantitative evaluation method and system for seismic data quality during mining based on signal correlation. By employing sensor array acquisition, time alignment, and anomaly classification based on the isolated forest algorithm, the evaluation efficiency and consistency are improved. By introducing workload parameters and equipment noise impact amplitude, the anomaly classification criteria are dynamically adjusted in two ways. This addresses the complex and variable working conditions during coal mining, enabling rigorous detection during high-intensity mining and optimizing computational resource allocation under strong noise interference, thus ensuring the accuracy of evaluation results and saving computational resources. Based on the principle of signal spatial correlation, it can accurately identify inconsistent signals caused by noise or sensor malfunctions, improving the accuracy of coal mining geological data analysis and providing key technical support for coal mine safety monitoring. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 A flowchart illustrating the automatic quantitative evaluation method for the quality of seismic data acquired during mining based on signal correlation, provided in an embodiment of this application. Figure 2 A schematic diagram of the structure of an automatic quantitative evaluation system for the quality of seismic data acquired during mining based on signal correlation, provided in an embodiment of this application; The components represented by each number in the attached diagram are explained below: Signal acquisition and alignment module 11, intensity parameter configuration module 12, noise standard correction module 13, and anomaly classification and evaluation module 14. Detailed Implementation

[0011] This application provides an automatic quantitative evaluation method and system for the quality of seismic data during mining based on signal correlation, which is used to address the technical problem of inaccurate analysis of coal mining geological data in the prior art.

[0012] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0013] It should be noted that the terms "comprising" and "having" are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to these processes, methods, products, or devices.

[0014] Example 1, as Figure 1 As shown, this application provides an automatic quantitative evaluation method for the quality of seismic data acquired during mining based on signal correlation. The method includes: S100: During the coal mining process, a sensor array is used to collect seismic signals from multiple locations within a preset time window, and time alignment processing is performed to obtain multiple seismic signal sequences. In this embodiment, during coal mining, a sensor array is used to collect seismic signals from multiple locations within a preset time window, and these signals are then time-aligned to obtain multiple seismic signal sequences. This step involves deploying multiple sensors within the mining area to collect seismic signals during the coal mining process, including vibration signals generated by different geological features and noise signals from the coal mining machine itself, such as vibrations from gears and belts. Through a standardized acquisition and processing procedure, multi-location, time-synchronized seismic signal sequences are obtained, providing fundamental data support for subsequent data quality evaluation.

[0015] Specifically, step S100 includes the following sub-steps: During the coal mining process, a sensor array is used to collect seismic signals from multiple locations within a preset time window. The test obtained the time delay of the seismic signals monitored at multiple locations during mining, and the time delay of multiple signals was obtained. The seismic signals during mining were then time-aligned. The time-aligned seismic signals were arranged in chronological order to obtain multiple seismic signal sequences.

[0016] In this embodiment, firstly, during coal mining, a sensor array is used to collect seismic signals from multiple locations within a preset time window. The seismic signals collected during the mining process are obtained by deploying a sensor array on the coal face and surrounding rock strata. Seismic signals from the mining process refer to vibration signals directly or indirectly triggered and collected during coal mining activities; they are a mixture of effective geological information and equipment noise. A sensor array is a collection of multiple sensors arranged in a specific geometric pattern for spatial measurement. It can simultaneously capture signals from multiple locations for locating seismic sources and analyzing wave field propagation. For example, 10 vibration sensors are evenly deployed within a 50-meter radius of the mining area. A preset time window of 30 minutes and a sampling interval of 5 seconds are set, meaning the amplitude and frequency of the signal are recorded every 5 seconds. A total of 30 minutes × (60 seconds / minute) / 5 seconds = 360 seismic signals from the mining process are collected within 30 minutes. Each seismic signal contains the seismic wave characteristics within that 5-second time period, such as an average voltage amplitude of 5V and a dominant frequency of 200Hz.

[0017] Secondly, the time delays of seismic signals monitored at multiple locations were obtained through testing, resulting in multiple signal delays. These seismic signals were then time-aligned. Because the distances from each sensor to the seismic source differ, and seismic waves require time to propagate, the arrival time of the same vibration signal at different sensors varies, resulting in signal delays. This step determines the time delay value of each sensor relative to a reference sensor through experiments or theoretical calculations. Then, the data sequence of each sensor is shifted forward or backward along the time axis by the corresponding time delay value, thus aligning the same vibration signal recorded by all sensors in time. For example, through synchronization testing, a vibration source at a known time is manually triggered, and the signal delays of each sensor are obtained, such as sensor A with a delay of 0.02 seconds and sensor B with a delay of 0.05 seconds. Then, all signals are time-calibrated, advancing the signal of sensor B by 0.03 seconds to synchronize it with sensor A.

[0018] Finally, the time-aligned seismic signals are arranged chronologically to obtain multiple seismic signal sequences. After time alignment, the calibrated data from each sensor are arranged chronologically to form an ordered dataset. For example, the 360 ​​signals from sensor A are ordered chronologically as [A1 (0 seconds), A2 (5 seconds), ..., A360 (1795 seconds)]. Similarly, signal sequences from sensors B, C, ... are generated, ultimately resulting in 10 sets of time-synchronized seismic signal sequences. Each sequence contains 360 data points arranged chronologically. At any given time point, the records from all 10 sensors can be acquired simultaneously, facilitating subsequent spatial correlation analysis.

[0019] In this embodiment, raw seismic signal data from different sensors and at different times are processed into a unified, structured signal sequence, which facilitates subsequent batch processing and comparative analysis. Through time alignment processing, it is ensured that the same vibration event recorded by sensors at different spatial locations is aligned on the time axis, laying the foundation for subsequent signal correlation analysis.

[0020] S200: Obtain the working intensity parameters of the coal mining equipment, and configure the first anomaly classification standard according to the working intensity parameters.

[0021] In this embodiment, the working intensity parameters of the coal mining equipment are obtained, and a first anomaly classification standard is configured based on these parameters. The working intensity of the coal mining equipment directly determines the quality of the seismic signal source during mining. For example, under high working intensity, such as during active coal cutting, the disturbance of the coal and rock mass by the mining equipment is strong and regular, generating ideal source signals with strong energy, high consistency, and good spatial correlation. At this time, the data signal-to-noise ratio is high, and a more stringent anomaly classification standard should be adopted to accurately identify any possible minor anomalies, thereby improving the accuracy of data analysis. However, under low working intensity, such as when the coal mining equipment is stopped, under maintenance, or running unloaded, the signals generated by the coal and rock mass are weak, irregular, or unrepresentative. At this time, the data signal-to-noise ratio is low. If a high standard is still used for anomaly detection, not only will computational resources be wasted, but most of the detected anomalies will also be meaningless noise, making the conclusions unreliable. This step, by introducing the working intensity parameter, can dynamically adjust the evaluation criteria according to the inherent quality conditions of the current signal source, achieving a reasonable allocation of computational resources and ensuring the effectiveness of the evaluation conclusions.

[0022] Specifically, step S200 includes the following sub-steps: Obtain preset anomaly classification criteria, which include a preset number of classifications, set according to the total number of seismic signals collected within the seismic signal sequence. Obtain the average working intensity parameter of the coal mining equipment, and calculate the ratio of the working intensity parameter to the average working intensity parameter as the working intensity coefficient; The workload coefficient is used to adjust and calculate the preset anomaly classification standard to obtain the first anomaly classification standard.

[0023] In this embodiment, firstly, a preset anomaly classification standard is obtained. This standard includes a preset number of classifications, which is set based on the total number of seismic signals collected within the seismic signal sequence. An initial, neutral anomaly classification standard is needed as a benchmark for adjustment. A key parameter in the preset anomaly classification standard is the preset number of classifications, which defines how many groups of signals are randomly selected from the signal sequence for anomaly analysis. This number is typically set as a certain proportion of the total data volume to ensure statistical significance. For example, in the above process, if a sequence containing 360 seismic signals is obtained, and the preset number of classifications is half the total number of seismic signals collected within the sequence, then the preset number of classifications is 180, meaning 180 seismic signals are randomly selected for anomaly analysis.

[0024] Secondly, the average working intensity parameter of the coal mining equipment is obtained, and the ratio of this working intensity parameter to the average working intensity parameter is calculated as the working intensity coefficient. To quantify the current working state, real-time working intensity parameters are obtained, such as the current of the coal mining machine's cutting motor and the drum speed, and compared with a historical or theoretical average working intensity parameter. The ratio of the two is calculated to obtain the working intensity coefficient, which directly reflects the degree of deviation of the current intensity from the average level. For example, if the average working intensity parameter of the coal mining machine is an average current of 100A, and the currently monitored working intensity parameter is 150A, then the calculated working intensity coefficient = 150 / 100 = 1.5, indicating that the current working state is high-intensity.

[0025] Finally, the preset anomaly classification criteria are adjusted using the aforementioned workload coefficient to obtain the first anomaly classification criterion. The preset baseline is adjusted using the workload coefficient obtained in the previous step to generate a first anomaly classification criterion adapted to the current situation. Specifically, the workload coefficient is multiplied by the preset number of classifications, and the result is rounded down, ensuring that the result is within a valid range, with a minimum of 1 and a maximum not exceeding the total number of signals. For example, if the preset number of classifications is 180 and the workload coefficient is 1.5, the adjusted number of classifications = 1.5 × 180 = 270. Since 270 is less than the total number of signals 360 and greater than 1, the number of classifications in the first anomaly classification criterion is set to 270, indicating that more signal groups need to be extracted and analyzed, and more stringent checks need to be performed.

[0026] In this embodiment, by introducing a workload parameter, more computing power is invested in fine analysis under high workload conditions to improve data quality; and computing power is saved under low workload conditions, avoiding resource waste and preventing unreliable detection under poor signal quality conditions, thereby improving the accuracy and credibility of the final data quality evaluation results.

[0027] S300: Obtain the influence amplitude of coal mining equipment noise under the working intensity parameter and the seismic signal, adjust the first anomaly classification standard, and obtain the second anomaly classification standard.

[0028] In this embodiment, the influence amplitude of coal mining equipment noise within the seismic signal under the working intensity parameter is obtained, and the first anomaly classification standard is adjusted to obtain a second anomaly classification standard. In the above process, the anomaly classification standard has been initially adjusted according to the working intensity of the coal mining equipment, but the influence of the equipment's own noise on the seismic signal, such as gear vibration and no-load operation noise, is not considered. For example, when the coal mining machine is running no-load, the noise signal intensity may mask the actual geological vibration signal, leading to distorted anomaly classification results. By quantifying the noise influence amplitude, the first anomaly classification standard is optimized a second time, allowing the standard to dynamically adapt to the current noise level, avoiding over-detection of low signal-to-noise ratio data, and improving computational efficiency and evaluation accuracy.

[0029] Specifically, step S300 includes the following sub-steps: Based on the influence amplitude of coal mining equipment noise within the seismic signal under the aforementioned working intensity parameters, anomaly classification adjustment coefficients are obtained. The first anomaly classification standard is adjusted and calculated using the aforementioned anomaly classification adjustment coefficient to obtain the second anomaly classification standard.

[0030] Among them, based on the influence amplitude of coal mining equipment noise within the seismic signal under the aforementioned working intensity parameters, anomaly classification adjustment coefficients are obtained, including: The noise signal intensity of the coal mining equipment under no-load operation was obtained through testing under the stated working intensity parameters. Obtain the average signal intensity within the multiple seismic signal sequences obtained during seismic sampling; Calculate the ratio of the device noise signal strength to the average signal strength to obtain the anomaly classification adjustment coefficient.

[0031] In this embodiment, an anomaly classification adjustment coefficient is obtained based on the impact amplitude of coal mining equipment noise within the seismic signals during mining under the stated working intensity parameters. First, the equipment noise signal intensity of the coal mining equipment under no-load operation is tested and obtained under the stated working intensity parameters: Under the current working intensity parameters, the coal mining equipment is controlled to operate under no-load conditions, and the intensity of the pure equipment noise signal generated is measured through a sensor array. This intensity represents the baseline noise level under the current operating conditions. Second, the average signal intensity within the multiple seismic signal sequences during mining is obtained: From the multiple seismic signal sequences generated above, the average value of all signal amplitudes is extracted to reflect the overall energy level of the signal within the current time period. Finally, the ratio of the equipment noise signal intensity to the average signal intensity is calculated to obtain the anomaly classification adjustment coefficient: The ratio of the equipment noise signal intensity to the average signal intensity is used as the anomaly classification adjustment coefficient. The formula is: Anomaly Classification Adjustment Coefficient = Equipment Noise Signal Intensity / Average Intensity of Seismic Signal During Mining. This ratio directly quantifies the degree of noise influence in the total signal. The larger the proportion of noise signal intensity, the more serious the noise pollution, and the more relaxed the anomaly classification standard needs to be. For example, the noise signal intensity of the coal mining machine under no-load operation is measured to be 2.5V. From the multiple seismic signal sequences obtained above, the average intensity of all signals is statistically calculated to be 5V; the anomaly classification adjustment coefficient is then calculated as 2.5 / 5 = 0.5. The anomaly classification adjustment coefficient is generally ≤1.

[0032] Secondly, the anomaly classification adjustment coefficient is used to adjust the first anomaly classification standard to obtain the second anomaly classification standard. The calculation formula is: Second anomaly classification standard = (rounded) [First anomaly classification standard × Anomaly classification adjustment coefficient]. The lower the proportion of equipment noise, the better the signal quality, thus reducing the detection stringency to save computing power; the higher the proportion of equipment noise, the worse the signal quality, requiring maintaining or increasing the detection stringency and improving the classification quality inspection standard. For example, if the number of categories in the first anomaly classification standard is 270 and the anomaly classification adjustment coefficient is 0.5, then the number of categories in the second anomaly classification standard = 270 × 0.5 = 135. Therefore, only 135 data points need to be detected, eliminating the need to detect all data points, thus saving computing power.

[0033] In this embodiment, the shortcomings of relying solely on workload parameters are addressed by introducing noise impact amplitude as a secondary adjustment factor: associating with the idle noise of equipment to eliminate the bias of mechanical interference in the evaluation of seismic signals; dynamic resource allocation to reduce the detection scale to save computing power when the noise proportion is low and increase the detection rigor when the noise proportion is high; and combining the first and second anomaly classification standards to jointly improve the decision-making ability under complex coal mining conditions and ensure the accuracy of the final data.

[0034] S400: Using the second anomaly classification standard, the multiple seismic signal sequences obtained during mining are classified as anomalies to obtain anomaly coefficients, and the quality coefficients of the seismic data obtained during mining are calculated as evaluation results.

[0035] In this embodiment, the second anomaly classification standard is used to classify the multiple mining-related seismic signal sequences into anomalies, obtain anomaly coefficients, and calculate the mining-related seismic data quality coefficient as the evaluation result. In the complex environment of coal mines, effective signals and anomalous signals are often mixed. Introducing the Isolation Forest unsupervised machine learning algorithm can automatically learn the data distribution characteristics of normal signals, efficiently identify anomalous data that does not conform to the characteristics of the main signals, and finally output an objective data quality coefficient, providing a reliable basis for subsequent geological interpretation.

[0036] Specifically, step S400 includes the following sub-steps: According to the number of second categories within the second anomaly classification criteria, multiple sets of seismic signals are randomly selected from multiple seismic signal sequences. Each set of seismic signals includes seismic signals from multiple sensors at a given time. Obtain an abnormal signal classifier; Multiple sets of seismic signals are input into the anomaly signal classifier, and multiple anomaly classification results are output, wherein each anomaly classification result includes anomaly or normal. The proportion of anomalies within multiple anomaly classification results is calculated to obtain the anomaly coefficient, and the quality coefficient of the seismic data obtained during mining is calculated as the evaluation result.

[0037] In this embodiment, firstly, according to the second classification number within the second anomaly classification standard, multiple sets of follow-up seismic signal groups are randomly selected from multiple follow-up seismic signal sequences. Each follow-up seismic signal group includes follow-up seismic signals from multiple sensors at a given time. Based on the classification number determined in the second anomaly classification standard, a corresponding number of time points are randomly selected from the multiple aligned signal sequences obtained above. For each selected time point, the recorded data from all sensors at that time are extracted to form a follow-up seismic signal group, such that each group contains observations from multiple points in space at the same time. For example, with 10 sensors, a signal sequence length of 360, and a second classification number of 135, 135 signal combinations at the same time are randomly selected from the 10 sets of follow-up seismic signal sequences obtained above, forming 135 signal groups. Each group contains follow-up seismic signals from 10 sensors at the same time, such as Group 1: signals from sensors 1-10 at the 10th second; Group 2: signals from sensors 1-10 at the 20th second, and so on, ultimately resulting in 135 such data groups.

[0038] Secondly, an abnormal signal classifier is obtained, including: A set of sample seismic signals is acquired, where each sample seismic signal includes multiple types of seismic signal features, including amplitude and frequency. Based on these multiple types of seismic signal features, the set of sample seismic signals is divided into multiple sample seismic signal feature sets. For example, if 135 sample seismic signals are acquired, including normal geological signals, equipment noise signals, and anomalous geological signals, the amplitude and frequency features of each sample are extracted, and the sets are divided into amplitude feature sets and frequency feature sets.

[0039] Based on the isolated forest model, multiple anomaly feature classification trees are constructed using multiple sets of sample seismic signal features. Each anomaly feature classification tree includes multiple layers of signal feature classification nodes. Each layer of signal feature classification nodes performs binary classification on the input seismic signal features and inputs the binary classification results into the upper layer signal feature classification nodes for further classification.

[0040] Isolation Forest is a tree-structured anomaly detection algorithm that isolates a few anomalous data points into individual data points, while normal data forms dense clusters and anomalous data points into anomalous data points. The anomaly feature classification tree uses multiple layers of binary classification nodes based on a single signal feature to achieve layer-by-layer data separation. Isolation Forest is trained using historical samples, building multiple classification trees based on amplitude and frequency features. For example, an amplitude classification tree is constructed using an amplitude feature set, containing multiple layers of nodes: such as the first layer dividing by 0.5V (≥0.5V vs. <0.5V); the second layer dividing by 1.2V, and so on; similarly, a frequency classification tree is constructed using a frequency feature set, containing multiple layers of nodes: such as the first layer dividing by 100Hz (≥100Hz vs. <100Hz); the second layer dividing by 250Hz, and so on; and so on, forming multiple layers of nodes until all data points are isolated or the tree depth limit is reached, for example, building 100 layers of binary classification nodes.

[0041] An anomalous signal classifier is obtained based on multiple anomaly feature classification trees. The amplitude and frequency classification trees are integrated to form the anomalous signal classifier. If the frequency acquired by a sensor differs significantly from the frequencies of other sensors, it is likely to be classified as a single data point, thus indicating an anomaly. Conversely, if no single data point is identified in the currently monitored seismic signal feature group, it is considered normal, indicating high data quality and similar features acquired by multiple sensors. For example, in 10 signal groups, if the frequencies of 9 sensors are distributed between 95-105Hz, while one sensor has a frequency of 200Hz, this signal is isolated in the classification tree and determined to be an anomaly.

[0042] Secondly, multiple sets of seismic signals collected during mining are input into the anomaly signal classifier, and multiple anomaly classification results are output, including: Based on the characteristics of various seismic signals, the first seismic signal group within multiple seismic signal groups is divided to obtain multiple first seismic signal feature groups. The first seismic signal group, i.e., the data from all sensors at the same time, is taken and decomposed into different feature sets according to signal characteristics. For example, a seismic signal group containing data from 10 sensors: First, amplitude values ​​are extracted from each sensor's data to form an amplitude feature group containing 10 amplitude values, resulting in first seismic signal feature group A: [1.2, 1.1, 1.0, 1.3, 1.1, 1.2, 1.0, 1.1, 1.2, 1.0]; then, frequency values ​​are extracted to form a frequency feature group containing 10 frequency values, resulting in first seismic signal feature group B: [98, 101, 100, 99, 102, 101, 100, 200, 99, 100].

[0043] Multiple first seismic signal feature groups are input into multiple anomaly feature classification trees within the anomaly signal classifier for multi-level classification to obtain a first anomaly classification result. If any first seismic signal feature is classified as a single data point, the first anomaly classification result is considered anomaly; if no first seismic signal feature is classified as a single data point, the first anomaly classification result is considered normal. The separated feature groups are then processed by their corresponding anomaly feature classification trees within the anomaly signal classifier. For example, if the first seismic signal feature group A is input into the amplitude classification tree, the amplitude values ​​of each sensor are close, ranging from 1.0 to 1.3, forming a dense cluster with no isolated points, thus it is considered normal. If the first seismic signal feature group B is input into the frequency classification tree, most frequencies are concentrated in the 98–102 Hz range, but one sensor has a frequency of 200 Hz, far from the normal cluster; this 200 Hz point is isolated and considered anomaly.

[0044] Continue by inputting multiple sets of seismic signals from the mining process into the anomaly classifier, and obtain multiple anomaly classification results. Repeat the above steps, inputting all remaining sets of seismic signals from the mining process into the anomaly classifier in sequence, thereby obtaining the anomaly classification result corresponding to each signal set.

[0045] Finally, the proportion of anomalies within multiple anomaly classification results is calculated to obtain the anomaly coefficient, and the on-site seismic data quality coefficient is calculated as the evaluation result. For example, if 27 out of 270 data sets are identified as anomalies, then the anomaly coefficient = 27 / 270 = 0.1, and the overall signal quality coefficient = 1 - anomaly coefficient = 0.9, indicating that the on-site seismic data quality is 90%.

[0046] In this embodiment, the data is comprehensively examined from multiple key feature dimensions. Any problem in any dimension will be captured, which improves the sensitivity and reliability of anomaly identification. It avoids the waste of computing power in full signal detection and reduces misjudgment through multi-feature classification. The fully automated classification and calculation process enables efficient and objective quality screening of massive amounts of data, improves evaluation efficiency, and enhances the accuracy of coal mining geological data analysis.

[0047] Example 2, as Figure 2 As shown, this application provides an automatic quantitative evaluation system for the quality of seismic data acquired during mining based on signal correlation, including: The signal acquisition and alignment module 11 is used to acquire seismic signals from multiple locations within a preset time window during the coal mining process through a sensor array, and perform time alignment processing to obtain multiple seismic signal sequences. The intensity parameter configuration module 12 is used to obtain the working intensity parameters of the coal mining equipment and configure the first anomaly classification standard according to the working intensity parameters. The noise standard correction module 13 is used to obtain the influence amplitude of coal mining equipment noise under the working intensity parameter and the seismic signal during mining, and to adjust the first anomaly classification standard to obtain the second anomaly classification standard. The anomaly classification and evaluation module 14 is used to classify the multiple seismic signal sequences obtained during mining using the second anomaly classification standard, obtain anomaly coefficients, and calculate the seismic data quality coefficients obtained during mining as evaluation results.

[0048] In one embodiment, the signal acquisition and alignment module 11 is further configured to: During the coal mining process, a sensor array is used to collect seismic signals from multiple locations within a preset time window. The test obtained the time delay of the seismic signals monitored at multiple locations during mining, and the time delay of multiple signals was obtained. The seismic signals during mining were then time-aligned. The time-aligned seismic signals were arranged in chronological order to obtain multiple seismic signal sequences.

[0049] In one embodiment, the strength parameter configuration module 12 is further configured to: Obtain preset anomaly classification criteria, which include a preset number of classifications, set according to the total number of seismic signals collected within the seismic signal sequence. Obtain the average working intensity parameter of the coal mining equipment, and calculate the ratio of the working intensity parameter to the average working intensity parameter as the working intensity coefficient; The workload coefficient is used to adjust and calculate the preset anomaly classification standard to obtain the first anomaly classification standard.

[0050] In one embodiment, the noise standard correction module 13 is further configured to: Based on the influence amplitude of coal mining equipment noise within the seismic signal under the aforementioned working intensity parameters, anomaly classification adjustment coefficients are obtained. The first anomaly classification standard is adjusted and calculated using the aforementioned anomaly classification adjustment coefficient to obtain the second anomaly classification standard.

[0051] Among them, based on the influence amplitude of coal mining equipment noise within the seismic signal under the aforementioned working intensity parameters, anomaly classification adjustment coefficients are obtained, including: The noise signal intensity of the coal mining equipment under no-load operation was obtained through testing under the stated working intensity parameters. Obtain the average signal intensity within the multiple seismic signal sequences obtained during seismic sampling; Calculate the ratio of the device noise signal strength to the average signal strength to obtain the anomaly classification adjustment coefficient.

[0052] In one embodiment, the anomaly classification and evaluation module 14 is further configured to: According to the number of second categories within the second anomaly classification criteria, multiple sets of seismic signals are randomly selected from multiple seismic signal sequences. Each set of seismic signals includes seismic signals from multiple sensors at a given time. Obtain an abnormal signal classifier; Multiple sets of seismic signals are input into the anomaly signal classifier, and multiple anomaly classification results are output, wherein each anomaly classification result includes anomaly or normal. The proportion of anomalies within multiple anomaly classification results is calculated to obtain the anomaly coefficient, and the quality coefficient of the seismic data obtained during mining is calculated as the evaluation result.

[0053] The abnormal signal classifier includes: Obtain a set of sample seismic signals, where each sample seismic signal includes multiple types of seismic signal features, including amplitude and frequency; Based on the characteristics of multiple types of seismic signals, the sample seismic signal set is divided to obtain multiple sample seismic signal feature sets; Based on the isolated forest model and multiple seismic signal features, multiple sample seismic signal feature sets are used to construct multiple anomaly feature classification trees. Each anomaly feature classification tree includes multiple layers of signal feature classification nodes. Each layer of signal feature classification nodes performs binary classification on the input seismic signal features and inputs the binary classification results into the upper layer of signal feature classification nodes for further classification. An abnormal signal classifier is obtained based on multiple abnormal feature classification trees.

[0054] The process involves inputting multiple sets of seismic signals obtained during mining into the anomaly signal classifier, which outputs multiple anomaly classification results, including: Based on the characteristics of multiple types of seismic signals, the first seismic signal group within the multiple seismic signal groups is divided to obtain multiple first seismic signal characteristic groups. Multiple groups of first seismic signal features are respectively input into multiple anomaly feature classification trees in the anomaly signal classifier to perform multi-level classification and obtain the first anomaly classification result. If any first seismic signal feature is classified as a single data, the first anomaly classification result is anomaly; if no first seismic signal feature is classified as a single data, the first anomaly classification result is normal. Multiple sets of seismic signals were input into the anomaly signal classifier, and the classification output yielded multiple anomaly classification results.

[0055] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0056] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0057] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. An automatic quantitative evaluation method for the quality of seismic data acquired during seismic mining based on signal correlation, characterized in that, The method includes: During the coal mining process, a sensor array is used to collect seismic signals from multiple locations within a preset time window, and time alignment processing is performed to obtain multiple seismic signal sequences. Obtain the working intensity parameters of the coal mining equipment, and configure a first anomaly classification standard based on the working intensity parameters; The influence amplitude of coal mining equipment noise under the working intensity parameter and the seismic signal is obtained. The first anomaly classification standard is adjusted to obtain the second anomaly classification standard. The second anomaly classification standard is used to classify the multiple seismic signal sequences obtained during mining, obtain anomaly coefficients, and calculate the seismic data quality coefficients obtained during mining as evaluation results.

2. The automatic quantitative evaluation method for the quality of seismic data acquired during mining based on signal correlation as described in claim 1, characterized in that, During coal mining, a sensor array is used to collect seismic signals from multiple locations within a preset time window. These signals are then time-aligned to obtain multiple seismic signal sequences, including: During the coal mining process, a sensor array is used to collect seismic signals from multiple locations within a preset time window. The test obtained the time delay of the seismic signals monitored at multiple locations during mining, and the time delay of multiple signals was obtained. The seismic signals during mining were then time-aligned. The time-aligned seismic signals were arranged in chronological order to obtain multiple seismic signal sequences.

3. The automatic quantitative evaluation method for the quality of seismic data acquired during mining based on signal correlation as described in claim 1, characterized in that, Obtain the working intensity parameters of the coal mining equipment, and configure a first anomaly classification standard based on the working intensity parameters, including: Obtain preset anomaly classification criteria, which include a preset number of classifications, set according to the total number of seismic signals collected within the seismic signal sequence. Obtain the average working intensity parameter of the coal mining equipment, and calculate the ratio of the working intensity parameter to the average working intensity parameter as the working intensity coefficient; The workload coefficient is used to adjust and calculate the preset anomaly classification standard to obtain the first anomaly classification standard.

4. The automatic quantitative evaluation method for the quality of seismic data acquired during mining based on signal correlation as described in claim 1, characterized in that, The influence amplitude of coal mining equipment noise within the seismic signal during mining is obtained under the aforementioned working intensity parameters. The first anomaly classification standard is adjusted to obtain a second anomaly classification standard, including: Based on the influence amplitude of coal mining equipment noise within the seismic signal under the aforementioned working intensity parameters, anomaly classification adjustment coefficients are obtained. The first anomaly classification standard is adjusted and calculated using the aforementioned anomaly classification adjustment coefficient to obtain the second anomaly classification standard.

5. The automatic quantitative evaluation method for the quality of seismic data acquired during mining based on signal correlation as described in claim 4, characterized in that, Based on the impact amplitude of coal mining equipment noise within the seismic signal under the aforementioned working intensity parameters, anomaly classification adjustment coefficients are obtained, including: The noise signal intensity of the coal mining equipment under no-load operation was obtained through testing under the stated working intensity parameters. Obtain the average signal intensity within the multiple seismic signal sequences obtained during seismic sampling; Calculate the ratio of the device noise signal strength to the average signal strength to obtain the anomaly classification adjustment coefficient.

6. The automatic quantitative evaluation method for the quality of seismic data acquired during mining based on signal correlation as described in claim 1, characterized in that, The multiple seismic signal sequences acquired during mining are classified for anomalies using the second anomaly classification standard to obtain anomaly coefficients. The quality coefficients of the acquired seismic data are then calculated as evaluation results, including: According to the number of second categories within the second anomaly classification criteria, multiple sets of seismic signals are randomly selected from multiple seismic signal sequences. Each set of seismic signals includes seismic signals from multiple sensors at a given time. Obtain an abnormal signal classifier; Multiple sets of seismic signals are input into the anomaly signal classifier, and multiple anomaly classification results are output, wherein each anomaly classification result includes anomaly or normal. The proportion of anomalies within multiple anomaly classification results is calculated to obtain the anomaly coefficient, and the quality coefficient of the seismic data obtained during mining is calculated as the evaluation result.

7. The automatic quantitative evaluation method for the quality of seismic data acquired during mining based on signal correlation as described in claim 6, characterized in that, Obtain an anomaly signal classifier, including: Obtain a set of sample seismic signals, where each sample seismic signal includes multiple types of seismic signal features, including amplitude and frequency; Based on the characteristics of multiple types of seismic signals, the sample seismic signal set is divided to obtain multiple sample seismic signal feature sets; Based on the isolated forest model and multiple seismic signal features, multiple sample seismic signal feature sets are used to construct multiple anomaly feature classification trees. Each anomaly feature classification tree includes multiple layers of signal feature classification nodes. Each layer of signal feature classification nodes performs binary classification on the input seismic signal features and inputs the binary classification results into the upper layer of signal feature classification nodes for further classification. An abnormal signal classifier is obtained based on multiple abnormal feature classification trees.

8. The automatic quantitative evaluation method for the quality of seismic data acquired during mining based on signal correlation as described in claim 7, characterized in that, Multiple sets of seismic signals collected during mining are input into the anomaly signal classifier, which outputs multiple anomaly classification results, including: Based on the characteristics of multiple types of seismic signals, the first seismic signal group within the multiple seismic signal groups is divided to obtain multiple first seismic signal characteristic groups. Multiple groups of first seismic signal features are respectively input into multiple anomaly feature classification trees in the anomaly signal classifier to perform multi-level classification and obtain the first anomaly classification result. If any first seismic signal feature is classified as a single data, the first anomaly classification result is anomaly; if no first seismic signal feature is classified as a single data, the first anomaly classification result is normal. Multiple sets of seismic signals were input into the anomaly signal classifier, and the classification output yielded multiple anomaly classification results.

9. An automatic quantitative evaluation system for the quality of seismic data acquired during seismic mining based on signal correlation, characterized in that: The system is used to implement the automatic quantitative evaluation method for the quality of seismic data acquired during mining based on signal correlation as described in any one of claims 1-8, the system comprising: The signal acquisition and alignment module is used to acquire seismic signals from multiple locations within a preset time window during coal mining through a sensor array, and perform time alignment processing to obtain multiple seismic signal sequences. The intensity parameter configuration module is used to obtain the working intensity parameters of the coal mining equipment and configure the first anomaly classification standard according to the working intensity parameters; The noise standard correction module is used to obtain the influence amplitude of coal mining equipment noise under the working intensity parameter and the seismic signal during mining, and to adjust the first anomaly classification standard to obtain the second anomaly classification standard. The anomaly classification and evaluation module is used to classify the multiple seismic signal sequences obtained during mining using the second anomaly classification standard, obtain anomaly coefficients, and calculate the seismic data quality coefficients obtained during mining as evaluation results.

Images